A strategy of LNG is to test candidate alleles against intermediate phenotypes including alcohol response and benzodiazepine response, which are predictive of vulnerability. Datasets for these two intermediate phenotypes have been developed in collaboration with their acknowledged experts: M. Schuckit who has prospectively followed a cohort of 450 men and has available psychophysiological and subjective measures of response as well as clinical outcome, and D. Cowley, who had collected sensitive and accurate eye-movement measures following benzodiazepine challenge [using the paradigm originally developed and brought to Seattle by D. Hommer, now Acting Chief of LCS]. Thus far we have reported preliminary, but positive results with a common, non-conservative GABAA alpha6 amino acid substitution discovered by LNG as well as with the functional serotonin transporter polymorphism HTTLPR described by P. Lesch [Schuckit et al; Iwata et al]. The linkage to HTTLPR was improved by use of a new functional allele we discovered at this locus. The GABAA alpha6 gene represents an example of where we have been able to find convergence between a rodent genetic finding and the human. Korpi et al detected a GABAA alpha6 amino acid substitution altering the sensitivity of this receptor to alcohol and benzodiazepines and this substitution predicts the difference in alcohol sensitivity observed between alcohol accepting and non-accepting Finnish rats. The GABAA clusters on human chromosomes 4 and 5 are implicated by linkage studies in the human [LNG, Long et al] and in the rodent [Portland group], respectively. Nevertheless, except for the GABAA alpha6 result in the rat, alpha6 might not have been the first GABAA receptor we evaluated in the human because the GABAA alpha6 is restricted in its expression to cerebellum. As discussed above, LNG was able to detect a common, non-conservative amino acid substitution Pro385Ser and obtain preliminary evidence for involvement in alcohol and benzodiazepine sensitivity The alcohol response dataset is being expanded according to M.Schuckit's resources to collect these individuals. We now have convergent data from two human isolates for a role for GABAA receptors in alcoholism. Using a six-locus haplotype for the Chr 5 GABAA gene cluster, we detected linkage and linkage disequilibrium in alcoholism in both Finns and SW American Indians. TrimHap (MacLean et al, 2000) was used to establish the most likely location of a gene influencing alcoholism. In both samples, this location was in the alpha6 or in the region of alpha6 and the adjacent gene for the alpha2 subunit, a subunit which is necessary for modulation of GABAA receptor function by ethanol. This work has recently been expanded with multi-locus dense SNP maps for both the chr 5 cluster and chr 4 clusters. Pharmacogenetics of treatment response. A radical change in alcoholism research is that several pharmacological treatments have recently become available that improve the ability of alcoholics to maintain abstinence. The previously available drug - disulfiram - had a direct counterpart in the ALDH2 Glu487Lys variant which created a natural blockade of the same metabolic enzyme. The new drugs, particularly sertraline and naltrexone, could also point directly to molecular targets which could be either sources of variation in treatment response or sources of differential vulnerability. Two principal gene targets are the serotonin transporter [sertraline] and the m opioid receptor [naltrexone], and other gene targets are also directly inferred: HTR1A and HTR1B autoreceptors, postsynaptic serotonin receptors, opioid propeptides, and various genes involved in the function of interacting neurotransmitters such as dopamine. LNG is focusing efforts in sequence variant detection towards these genes and by initiating a major collaboration led by Dr. S. O'Malley towards the pharmacogenetics of naltrexone and sertralinne response among alcoholics in a NW Indian tribe and towards the pharmacogenetics of naltrexone and acamprosate in the COMBINE multicenter study. The human research protocol for the O'Malley study is now approved by the NIAAA IRB and this study is underway. LNG's molecular studies have already yielded a common, nonconservative mu opioid receptor variant which M. Kreek and colleagues have reported alters affinity of the receptor for endomorphin. Effect of this polylmorphism on phenotype is now being studied in large cohorts of alcoholics and opioid addicts. Ten Tribes Study. LNG is a leading exponent of gene/environment interactions in alcoholism, including the use of cross-population studies. The Ten Tribes Study is a gene/environment interaction study encompassing the collection of the first large multi-population dataset specifically suited for evaluation of gene/environment interaction: the Ten Tribes Study. This study, which is approximately 2/3 complete involves the collection of DNA and psychiatric interview [AUDADIS] data on 300 demographically sampled individuals from 10 different American Indian tribes, including tribes with widely disparate rates of alcoholism and alcohol associated problems. This dataset will be used to study the effects of social and historical determinants on alcoholism, effects of alcoholism on communities [especially rates and types of trauma], and interaction of genetic factors with environmental loadings and thresholds.